Welcome back to Invested Early.

A Bitcoin fintech wiped out its secured debt this week and spiked 160% intraday. Iran tensions sent an oil stock flying 57%. And a $7 power company building AI infrastructure in Texas just caught the attention of OpenAI speculators. We're covering all of it, plus a question that separates tourists from serious small-cap investors: how do you actually read the financials of a company that has no earnings yet?

Here's what's inside today:

📊 How to read small-cap financials when there are no earnings to read
⚡ The $7 stock building a 17-gigawatt AI power campus in Texas
🤖 AltIndex put three of its strategies on Autopilot
📈 Top gainers (and biggest losers) this week

Let’s get into it!

This is not financial advice. Always do your own research. Past performance doesn't guarantee future results.

Most small caps don't have earnings. A lot of them don't have revenue either. So what are you supposed to look at?

Cash runway is everything. Find the quarterly cash burn (operating cash flow, usually negative) and divide it into total cash on hand. That tells you how many quarters until the company needs to raise money. Under four quarters? Expect dilution coming. Companies that raise capital before they actually need it tend to get better terms and avoid desperation dilution.

Watch the share count. Pull up the balance sheet from a year ago and compare it to today. If shares outstanding grew by 30% while the stock price stayed flat, existing shareholders got diluted by 30%. Some companies grow their share count faster than their business. That's a red flag.

Revenue trajectory beats revenue size. A company doing $5 million in revenue that grew 200% year-over-year is more interesting than one doing $50 million that grew 10%. At this stage, you're betting on trajectory.

Ignore GAAP earnings. Stock compensation, depreciation, one-time charges: they can make a growing company look like a disaster. Focus on cash flow and unit economics instead.

None of this is complicated. It's also easy to skip when a stock is already running and you just want in. That's usually when it matters most.

You've heard the thesis: AI needs compute, compute needs power, power is the bottleneck. This week the market reminded everyone just how real that bottleneck is.

The news: Fermi Inc. (FRMI) jumped 22.6% last week after reports that OpenAI is in talks to lease a 10-gigawatt data center in Ohio, potentially backed by Nvidia. The deal has nothing to do with Fermi directly, but investors connected the dots: if OpenAI is willing to commit to that kind of capacity, Fermi's Project Matador just got a lot more interesting.

Why it matters: Fermi is building what it calls a "behind-the-meter" AI campus in Carson County, Texas, designed to deliver power directly to data centers without relying on the traditional grid. The project spans 7,500 acres and targets 17 gigawatts at full scale using natural gas, nuclear, solar, and battery storage. The company says it has secured over 2 gigawatts of owned and contracted supply, plus permits for another 6 gigawatts of generation capacity.

The stock trades a bit above $7 currently, with investors betting that hyperscale AI buyers care more about "time-to-power" than anything else right now. Citizens analysts noted in a recent report that Fermi's permitted and potentially powered capacity by 2027 makes it one of the few plays positioned for this wave.

The honest risk: Fermi has zero revenue. Its most recent 10-Q filing shows no tenant agreements signed. The company lost $189 million last quarter (mostly non-cash), holds $243 million in cash against $421 million in debt, and is in the middle of a proxy fight with its co-founder. The OpenAI news was about Ohio, not Texas, and no Fermi lease is in hand. There's no signed lease here, just a bet that one might come. If you're playing it, keep the position small.

Our partners at AltIndex put three of their strategies on Autopilot this week. Instead of just showing you which stocks their alternative data signals are flagging, they’ve made it so you can set your capital to be automatically invested for you using AltIndex’s data. (Autopilot is a separate SEC-registered platform; Invested Inc. may earn a commission on referrals.)

The one that we’re most interested in? Reddit's Favorites. It screens for stocks between $1 billion and $100 billion in market cap that are getting unusual Reddit attention, then applies a quality filter: high AI score, growing revenue, positive cash flow. In their backtest, it returned a hypothetical 89.9% over 24 months, with the smoothest ride of the three strategies.

The other two, AltIndex AI Top 10 and Hyper-Hiring Growth, target different signals (overall AI score strength and aggressive LinkedIn job postings, respectively). All three rebalance monthly, and your money never leaves your own brokerage.

The catch: backtests are backtests, not live results. If you've been curious how to trade on alternative data without doing all the work yourself, this is the cleanest version we've seen.

Explore the portfolios →

Here's where yesterday’s biggest moves landed.

Yesterday’s top gainers were a chip-and-memory story, which fits this edition's main event. Western Digital (WDC) is up a jaw-dropping 880% over the past year on the AI memory boom, AXT (AXTI) substrates have run even harder, and Wolfspeed (WOLF) keeps grinding higher. On the other side, Fox Corporation fell double digits across both its share classes after it announced plans to acquire Roku (ROKU), education names Stride (LRN) and McGraw Hill (MH) slid, and Sasol (SSL) gave back ground as oil prices fell.

(Data: Yahoo Finance.)

This week was a nice little reminder of what makes small caps small caps: a Bitcoin fintech wipes out its debt and spikes 160% intraday, an oil stock rips 57% on geopolitical headlines, and a $7 power play with zero revenue catches a bid on OpenAI speculation. The upside is real, but so is the risk.

We'll be back soon. Watching a name we should cover? Hit reply and let us know.

Cheers,
— Brandon & Blake of Invested Inc.